# Roi Proof Generator

> |

- **Type:** Skill
- **Install:** `agentstack add skill-swan-gtm-gtm-skills-roi-proof-generator`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [swan-gtm](https://agentstack.voostack.com/s/swan-gtm)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [swan-gtm](https://github.com/swan-gtm)
- **Source:** https://github.com/swan-gtm/gtm-skills/tree/main/skills/manny-medina/roi-proof-generator

## Install

```sh
agentstack add skill-swan-gtm-gtm-skills-roi-proof-generator
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

Use when agent value must be proven to the people paying for it — before a renewal, at a QBR, or monthly as a standing receipt. Produces a value receipt: tasks completed, hours returned, cost avoided, ROI multiple.

## Get the right kind of metrics

Three reporting categories exist; only one closes renewals:

- **usage reporting** — tasks run, tokens consumed. Engineering metrics.
- **performance analytics** — uptime, error rate, latency. Reliability metrics.
- **ROI reporting** — time saved, cost avoided, revenue generated. Commercial metrics.

Enterprise renewals depend on the third category alone. If the input data is usage-shaped, convert it before writing anything.

## Establish the benchmark

The receipt is only credible if the "before" number came from the customer. Pull or ask for:

- what each task cost before automation (time per task, loaded hourly rate of whoever did it)
- which value units the customer already tracks and cares about — use theirs, never invent proxies
- the price paid for the agent over the period

If no benchmark was ever agreed, capture one now and flag it: value definition belongs before deployment, not after.

## Compute the receipt

For the period (monthly is the right cadence — continuous proof beats a scramble at renewal):

1. **Tasks resolved** — count of completed units, in the customer's language ("847 support tickets resolved")
2. **Hours returned** — tasks × human-equivalent time, minus any human touch remaining ("average resolution 4 minutes; human equivalent 2.8 hours per ticket; 396 hours returned to your team")
3. **Cost avoided** — hours returned × loaded rate, plus any hard costs displaced
4. **ROI multiple** — value delivered ÷ price paid for the period

Show the arithmetic. A buyer who can re-derive the number will defend it internally; a black-box multiple gets discounted.

## Deliver it

Write the receipt as one page the champion can forward: headline multiple, the three numbers above, a short trend line versus prior periods, and zero adjectives. If this account gets a receipt every month, the renewal stops being a defense — you stop defending your price and start reminding the customer what losing the product costs.

## What good looks like

A great receipt uses benchmarks the customer stated, units they already track, and arithmetic they can check — and it existed before anyone asked. A mediocre one appears the week before renewal, built on vendor-estimated hours and activity counts. The traps: shipping usage metrics as if they were value, inventing human-equivalent times without customer sign-off, and averaging across accounts when margins and value vary enormously per account — always compute per customer.

MUST separate customer-agreed benchmarks from assumptions, and label any assumption. NEVER fabricate a benchmark or a rate. NEVER lead with tasks-completed alone — activity is not value.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [swan-gtm](https://github.com/swan-gtm)
- **Source:** [swan-gtm/gtm-skills](https://github.com/swan-gtm/gtm-skills)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-swan-gtm-gtm-skills-roi-proof-generator
- Seller: https://agentstack.voostack.com/s/swan-gtm
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.